The proposed project is focused on a neglected area within mental health research - how patient gender, ethnicity, and age influence the discussion and treatment of mental health problems during primary care medical visits. This is an important area of research because many patients see only primary care physicians, not mental health specialists, for their psychosocial concerns (Greenley et al. 1987; Leaf et al., 1988). To gain a better understanding of how patient gender, ethnicity, and age influence how psychosocial issues are being discussed and treated in primary care settings, an existing data set will be used. The data set includes audio-tapes and transcripts of 427 outpatient visits as well as post-visit interviews with patients and physicians. The data set was originally collected during 1995 at the general medicine and family practice clinics of the University of New Mexico Health Sciences Center. The data set is extremely unique in that it is one of the only collections of audio-tapes of physician-patient interactions collected in the United States that includes Spanish as well as English speakers. The following coding tools will be developed: (1) patient expression of emotional symptoms and discussion of family, social, or personal life, (2) patient physical symptom expression, and (3) physician and patient discussion of antidepressants and minor tranquilizers. Each of the 427 physician-patient interactions will be coded using the developed instruments. The coded physician-patient interaction data will then be merged with the medical and pharmacy record data and the patient and physician interview data. Hierarchial modeling and random effects regression techniques will be used to analyze the data.